β-Hairpin loop of eukaryotic initiation factor 1 (eIF1) mediates 40 S ribosome binding to regulate initiator tRNA(Met) recruitment and accuracy of AUG selection in vivo

Abstract

Recognition of the translation initiation codon is thought to require dissociation of eIF1 from the 40 S ribosomal subunit, enabling irreversible GTP hydrolysis (Pi release) by the eIF2·GTP·Met-tRNAi ternary complex (TC), rearrangement of the 40 S subunit to a closed conformation incompatible with scanning, and stable binding of Met-tRNAi to the P site. The crystal structure of a Tetrahymena 40 S·eIF1 complex revealed several basic amino acids in eIF1 contacting 18 S rRNA, and we tested the prediction that their counterparts in yeast eIF1 are required to prevent premature eIF1 dissociation from scanning ribosomes at non-AUG triplets. Supporting this idea, substituting Lys-60 in helix α1, or either Lys-37 or Arg-33 in β-hairpin loop-1, impairs binding of yeast eIF1 to 40 S·eIF1A complexes in vitro, and it confers increased initiation at UUG codons (Sui(-) phenotype) or lethality, in a manner suppressed by overexpressing the mutant proteins or by an eIF1A mutation (17-21) known to impede eIF1 dissociation in vitro. The eIF1 Sui(-) mutations also derepress translation of GCN4 mRNA, indicating impaired ternary complex loading, and this Gcd(-) phenotype is likewise suppressed by eIF1 overexpression or the 17-21 mutation. These findings indicate that direct contacts of eIF1 with 18 S rRNA seen in the Tetrahymena 40 S·eIF1 complex are crucial in yeast to stabilize the open conformation of the 40 S subunit and are required for rapid TC loading and ribosomal scanning and to impede rearrangement to the closed complex at non-AUG codons. Finally, we implicate the unstructured N-terminal tail of eIF1 in blocking rearrangement to the closed conformation in the scanning preinitiation complex.

Keywords: AUG Recognition; Ribosome Function; Ribosomes; Transfer RNA (tRNA); Translation; Translation Initiation Factors; eIF1; eIF2.

FIGURE 1.

A, hypothetical model describing the roles of eIF1 and eIF1A in start codon recognition. eIF1 is depicted in olive green, eIF1A in orange with the CTT in green (harboring the SE elements as blue spheres), and the helical domain and NTT (harboring the SI elements) shown in red. i, eIF1 and the scanning enhancer elements in the CTT of eIF1A stabilize an open conformation of the 40 S subunit that rapidly loads the TC. ii, following attachment to mRNA and eIF5, the 43 S PIC in the open conformation scans the mRNA for the start codon with Met-tRNA_i anchored in the P_OUT state. GAP domain in the eIF5-NTD (5N) stimulates GTP hydrolysis by the TC to produce an equilibrium between GTP and GDP·P_i, but release of P_i is impeded. iii, on AUG recognition, the Met-tRNA_i moves from the P_OUT to P_IN state, and eIF1 dissociates from its 40 S-binding site, and the eIF1A SE elements are displaced from the P site. eIF5-NTD dissociates from eIF2 and moves toward the eIF1A CTT, facilitating P_i release. The 48 S complex reflects the stable PIC produced after eIF1 dissociation and P_i release. eIF1 and the eIF1A SE elements promote P_OUT and impede the transition to the P_IN state, whereas the SI elements in the NTT and helical domain of eIF1A stabilize the P_IN state. (Adapted from Ref. , .) B, hypothetical effects on TC loading and start codon recognition of eIF1 mutations that weaken its binding to the 40 S subunit. i, mutations in helix α1 and loop-1 of eIF1 are predicted to impair its binding to the 40 S subunits and destabilize the P_OUT state, reducing the rate of TC loading and thereby conferring the Gcd⁻ phenotype. ii and iii, once the TC is bound to the PIC, these eIF1 mutations would also allow inappropriate dissociation of eIF1 from the 40 S subunit and attendant transition to the P_IN state to occur more readily at UUG codons, producing the Sui⁻ phenotype. Mutations in the eIF1A SE elements also destabilize the P_OUT state conferring Gcd⁻ and Sui⁻ phenotypes, whereas mutations in eIF1A SI elements suppress the effect of such eIF1A SE mutations, rescuing rapid TC binding to P_OUT and suppressing UUG initiation, by stabilizing the P_OUT state (Ssu⁻ phenotype). By enhancing 40 S binding of eIF1, SI mutations in eIF1A are predicted to mitigate the Gcd⁻ and Sui⁻ phenotypes of helix α1 and loop-1 mutations that weaken eIF1 binding to the 40 S subunit.

FIGURE 2.

A, alignment of eIF1 sequences from diverse eukaryotes. Secondary structures are indicated as a (α-helix) and b (β-sheet). Residues substituted by site-directed or random mutagenesis are indicated above, with green and purple designating Sui⁻ and lethal phenotypes, respectively, and black indicating no detectable Slg⁻ or Sui⁻ phenotypes. B and C, ribbon depiction of T. thermophila eIF1 bound to the 40 S subunit (PDB file 2XZM). 18 S rRNA phosphate backbone is shown in yellow with bases predicted to contact eIF1 residues in blue. Highlighted are residues in yeast eIF1 that when substituted confer Sui⁻ (in green) or lethal (in purple) phenotypes, and in parentheses are the corresponding residues in Tetrahymena.

FIGURE 3.

Substitutions of lysines 56, 60, and 59 in helix α1 confer Sui⁻ or lethal phenotypes. A, Slg⁻ and His⁺/Sui⁻ phenotypes of Lys-56 and Lys-60 substitutions. 10-Fold serial dilutions of derivatives of sui1Δ his4-301 strain JCY03 with the indicated SUI1 alleles on sc plasmids were spotted on synthetic complete medium lacking leucine (SC-Leu) supplemented with 0.3 mm histidine (+His) and on SC-Leu plus 0.003 mm His (−His) media and cultured at 30 °C for 2 or 7 days, respectively. B, quantification of Sui⁻ phenotypes. Strains from A also harboring plasmids p367 or p391 containing HIS4-lacZ reporters with an AUG or UUG start codon, respectively, were cultured in synthetic dextrose minimal medium (SD) supplemented with His and Trp at 30 °C to A₆₀₀ of ∼1.0, and β-galactosidase activities were measured in WCEs. The ratio of expression of the UUG to AUG reporter was calculated for replicate experiments, and the mean and S.E. were plotted. Numbers in parentheses are the means normalized to the WT value. C, derivatives of sui1Δ strain JCY03 containing the indicated SUI1 alleles were cultured in SD supplemented with His, Trp, and uracil at 30 °C to an A₆₀₀ of ∼1.0, and WCEs were subjected to Western analysis using antibodies against eIF1/Sui1 or eIF2Bϵ/Gcd6 (used as loading control). Two different amounts of each extract differing by a factor of 2 were loaded in successive lanes. Data in A–C on mutants K56E and K60E were published previously (16). D, JCY03 derivatives containing SUI1⁺ WT and a His₆-tagged version of the indicated SUI1 alleles (His-SUI1) were streaked on SC-Leu and supplemented with 5-FOA and incubated at 30 °C for 4 days. E, same strains as in D were cultured and subjected to Western analysis as in C using antibodies against His₆ epitope to detect His₆-eIF1 proteins.

FIGURE 4.

Overexpression of mutant eIF1 proteins suppresses their Sui⁻ or lethal phenotypes. A, JCY03 derivatives harboring the indicated SUI1 alleles in sc, low copy (I3N), or hc plasmids were analyzed on +His and −His media as in Fig. 3A. B, same strains as in A were cultured and subjected to Western analysis as in Fig. 3C. C, transformants of the strains from A containing the AUG or UUG HIS4-lacZ reporters were analyzed as in Fig. 3B. D, JCY03 derivatives containing SUI1⁺ and the indicated His-SUI1 alleles in hc plasmids were streaked on SC-Leu + 5-FOA and incubated at 30 °C for 5 days (d). E, same strains as in D were cultured and subjected to Western analysis as in Fig. 3E. F, JCY03 derivatives containing the indicated hc His-SUI1 alleles (after selection on SC-Leu + 5-FOA) were analyzed on +His and −His media for the indicated days as in Fig. 3A. G, transformants of the strains from F containing the AUG or UUG HIS4-lacZ reporters were analyzed as in Fig. 3B.

FIGURE 5.

A–C, tif11-17-21 Ssu⁻ mutation in the NTT of eIF1A partially suppresses the Sui⁻ phenotypes and the lethality of mutations in eIF1. A, 10-fold serial dilutions of derivatives of sui1Δ his4-301 P_GAL-TIF11 strain PMY03 containing plasmid-borne TIF11⁺ (pPMB76) or tif11-17-21 (pPMB77) with the indicated SUI1 alleles on sc or lc (I3N) plasmids were analyzed as in Fig. 3A except using SC-Leu-Trp instead of SC-Leu. B, strains from A containing the AUG or UUG HIS4-lacZ reporters were cultured in SD + His, and β-galactosidase activities were measured in WCEs. C, derivatives of sui1Δ his4-301 P_GAL-TIF11 strain PMY03 containing tif11-17-21 (pPMB77) with the indicated sc SUI1 alleles were streaked on SC-Leu-Trp supplemented with 5-FOA. D, derivatives of sui1Δ gcn2Δ strain CHY01 harboring the indicated SUI1 alleles in sc or lc (I3N) plasmids and either vector or hc TC plasmid p1780-IMT were cultured in SD + His, and β-galactosidase activities were measured in WCEs. E, derivatives of sui1Δ his4-301 strain JCY03 harboring the indicated SUI1 alleles in sc, lc (I3N), or hc plasmids and plasmid p180 were cultured in SD + His, Trp, and analyzed as in D. F, strains from A containing p180 were analyzed as in B.

FIGURE 6.

Substitutions of residues in β-hairpin loop-1 confer Sui⁻ or lethal phenotypes. A, derivatives of sui1Δ his4-301 strain JCY03 with the indicated sc SUI1 alleles were analyzed on +His and −His media and cultured at 30 or 37 °C for 2 days (+His) and at 30 °C for 7 days (−His). B, JCY03 derivatives containing SUI1⁺ and the indicated His-SUI1 alleles in sc or hc plasmids were streaked on SC-Leu + 5-FOA and incubated at 30 °C for 4 or 5 days. C, JCY03 derivatives containing SUI1⁺ and the indicated His -SUI1 alleles in sc plasmids were cultured and subjected to Western analysis as in Fig. 3E. D, same strains as in A were cultured and subjected to Western analysis as in Fig. 3C. E, strains from A containing the AUG or UUG HIS4-lacZ reporters were cultured and analyzed as in Fig. 3B. F, derivatives of strain JCY03 with the indicated hc SUI1 alleles analyzed on +His and −His media and cultured at 30 or 37 °C for the indicated days. G, strains from F containing the AUG or UUG HIS4-lacZ reporters were cultured and analyzed as in Fig. 3B.

FIGURE 7.

Overexpression of the eIF1 mutant protein and the tif11-17-21 mutation in eIF1A reduce the Slg⁻, Sui⁻, and Gcd⁻ phenotypes in SUI1 mutant cells. A, derivatives of sui1Δ his4-301 strain JCY03 with the indicated sc or hc SUI1 alleles were grown at 37 °C for 2 days on SC-Leu. B, derivatives of sui1Δ his4-301 P_GAL-TIF11 strain PMY03 containing TIF11⁺ (pPMB76) or tif11-17-21 (pPMB77) with the indicated sc SUI1 alleles were grown at 37 °C for 2 days on SC-Leu-Trp. C and D, strains from A (C) and from B (D) containing the AUG or UUG HIS4-lacZ reporters were cultured and analyzed as in Figs. 3B and 5B. E and F, strains from A (E) and from B (F) harboring plasmid p180 were cultured and analyzed as in Fig. 5, E and F.

FIGURE 8.

¹H-¹⁵N HSQC spectra of 0.2 mm¹⁵N-labeled WT (A), I3N (B), R33A (C), K37E (D), and K60E (E) eIF1proteins.

FIGURE 9.

Sui⁻ mutations in helix α1 and loop-1 impair binding to 40 S subunits in vitro. A, fluorescein-labeled WT or I3N mutant eIF1 protein was mixed with increasing concentrations of 40 S subunits in the presence or absence of 1 μm eIF1A, and the increase in fluorescence anisotropy was measured. B, binding of fluorescein-labeled WT or the indicated mutant eIF1 proteins to 40 S subunits in the presence of 1 μm eIF1A was measured as in A monitored using fluorescence anisotropy. C, K_d values from A and B calculated from fitting with hyperbolic curves. D, fluorescein-labeled WT eIF1 (5 nm) was pre-bound to 15 nm 40 S subunits in the presence of 1 μm eIF1A, mixed with increasing concentrations of unlabeled WT or mutant eIF1, and the change in anisotropy was measured. E, fluorescein-labeled WT eIF1 (15 nm) was mixed with 120 nm 40 S, 1 μm eIF1A, 150 nm pre-formed TC complex (300 nm eIF2, 150 nm tRNA_i, and 1 mm GDPNP), and 10 μm mRNA with UUG start codon, to form 43 S·mRNA (UUG) complexes. Increasing concentrations of unlabeled WT or mutant eIF1 were added, and the change in anisotropy was measured.

FIGURE 10.

I3N mutation in the NTT of eIF1 impairs TC recruitment and start codon selection in vitro. A, binding of TC to the 40 S subunits as a function of time measured by a native gel assay as the fraction of [³⁵S]Met-tRNA_i^Met associated with 40 S subunits in the presence of saturating amounts of eIF1 WT or mutant, eIF1A, and mRNA with an AUG start codon. Values are the averages of three independent experiments. B and C, kinetics of TC binding at several concentrations of 40 S subunit in the presence of mRNA with an AUG (B) or UUG (C) start codons was measured to obtain observed rate constants. The k_obs were plotted against the concentration of 40 S subunits and fit linearly to obtain the rate constant of TC binding to 40 S subunits. The k_on is the slope of the line. D and E, rate of TC dissociation from 43 S·mRNA complexes formed with [³⁵S]Met-tRNA_i^Met was determined by adding a chase of excess (≥300-fold) unlabeled TC, and the fraction of labeled TC bound to the 40 S subunits was monitored over time in native gels. Values are the averages of two or three independent experiments. F, surface representation of Tetrahymena eIF1 (shown in gray) bound to the 40 S subunit with the 18 S rRNA shown as a yellow surface and bases that contact eIF1 colored in blue (constructed from PDB file 2XZM). Highlighted in pale orange are residues corresponding to previously analyzed Sui⁻ substitutions, with the Tetrahymena residues listed in parentheses, including substitutions Q84P and D83G (13), G107R,G107K (15, 28), and the I93A,L96A,G97A substitutions in mutant 93–97 (5). Colored green at the 40 S interface are the basic residues in helix α1 and loop-1 analyzed here.